The goal for neuroprostheses is to restore neural function to a condition having the fidelity of a healthy system. However, contemporary neural prostheses, including cochlear implants, are not able to achieve this goal. The devices use electrical current to stimulate the neurons, which spreads in the tissue and consequently does not allow stimulation of focused populations of neurons. Therefore, high fidelity stimulation is no possible. In our model system, the cochlea, it has been argued that the performance of cochlear implant users could be increased significantly if more discrete locations of neurons situated along the electrode could be stimulated simultaneously. This might be possible with devices that use focal optical radiation to stimulate neurons. Today we know that infrared neural stimulation (INS) is possible, that stimulation rates can be achieved that allow encoding of acoustic information, that the spatial selectivity in the cochlea is about five times more selective than electrical stimulation, and that single channel stimulation in chronic experiments shows no functional damage of the cochlea over at least six weeks. The five-year project proposed here is a logical progression of our previous experiments. The aims include validating that the selectivity of INS will result in a larger number of independent channels, demonstrating that a three-channel device can safely stimulate an implanted cochlea over several weeks, and showing that each channel of multichannel INS can independently encode information to be perceived by the auditory system. At the conclusion of the project period we intend to present a prototype for a multi-channel neural interface for the human, here a cochlear implant. To determine the minimum channel separation for independent stimulation, we will implant a three-channel device in deaf cats. Recordings from the inferior colliculus will be used to construct spatial tuning curves (STCs). Non-overlapping STCs indicate separation of the channels. The distance between the stimulation sources will be altered systematically until independent stimulation at neighboring stimulation sources is obtained. By varying stimulus parameters such as the repetition rate, the pulse shape, and the delay between neighboring channels, the experiments will also provide information on the temporal properties of optical stimulation. Long-term stimulation after chronic implantation of a three-channel device into a cat cochlea will determine the safety. Evoked auditory responses will be measured and will provide information on cochlear function and safety. Results will be confirmed through histology. Measurements with temperature sensitive ink will provide important information on the heat load during stimulation. At the conclusion of this project, a prototype human optical cochlear implant will be constructed based on the physical and the optical requirements.